V.] SYMMETRICAL CORRESPONDENCES. 79
the upper part of its course, he says, follows a direction from west
to east1, he remarks that by some it was identified with a river
which had been discovered by the Nasamones far away in the
interior of Africa. He then adds, "reason favours that view, for
the Nile certainly flows out of Libya, dividing it down the middle,
and as I conceive, judging the unknown from the known, rises at
the same distance from its mouth as the Ister. This latter river
has its source in the country of the Celts near the city Pyrene,
and runs through the middle of Europe, dividing it into two
portions. The Celts live beyond the Pillars of Hercules, and
border on the Cynesians, who dwell at the extreme west of
Europe. Thus the Ister flows through the whole of Europe
before it finally empties itself into the Euxine at Istria, one of the
colonies of the Milesians2." With regard to the correspondence in
position of the mouths of the two rivers he further says:—" Egypt
lies almost exactly opposite the mountainous portion of Cilicia,
whence a lightly-equipped traveller may reach Sinope on the
Euxine in five days by the direct route. Sinope lies opposite the
place where the Ister falls into the sea. My opinion therefore is
that the Nile, as it traverses the whole of Libya, is of equal
length with the Ister3." In another passage he speaks of the Ister
as falling into the sea with its mouth facing the south-east4, and
from this it has sometimes been inferred that he intended to find
a resemblance between this change in the direction of its course
and the bend which he regarded the Nile as making from east to
north in the neighbourhood of Meroe; but perhaps we are hardly
justified in deducing so much from his words. In the passage
just quoted about the mouths of the rivers we may Attem te at
find an illustration of the more scientific side of Drawing a
Herodotus's mind; for in attempting to draw an Mcridlan-
imaginary line from Egypt to Cilicia, and thence by way of Sinope
to the Ister, he is evidently feeling his way towards a meridian of
longitude. Here the geographer's instinct for determining the
relative position of places, quickened by the map-maker's habit
of arranging them, is asserting itself. A similar rude essay in
mathematical geography is found in his describing the situation of
1 a. 31. « 3, 33.
* a. 34» 4 4- 99-